When your PC is attacked by 20 rats, it’s a bummer for the DM to make all those attack rolls. The obvious hack is to make one attack roll for the whole group of rats. That gives you pretty spiky results though: the only options are “All the rats hit” or “All the rats miss.”

The 5e DMG has a fix for this: a rule for making monster group (mob) attacks without rolling millions of attack rolls, and without a single all-or-nothing attack roll. You consult a chart which tells you how many attacks hit. For instance, if the monsters hit on a 13, 1/3 of the attackers hit.

I love this idea, and I suggested that it is complete enough to form a whole mass combat system, but after using it in play, I’ve found some problems I’d like to address.

1) There’s no d20 rolls at all. Mob combat is different from any other D&D task resolution.

2) It’s completely smooth. An AC 18 fighter being shot at by 20 hobgoblins is hit by a steady and predictable 5 arrows per round.

3) It requires a chart – not a big one but not one that’s easy to keep in your memory. Like the attack matrix charts in 1e, it’s a page you have to bookmark.

4) Because there are no die rolls, it doesn’t work naturally with advantage/disadvantage and crits are impossible.

Here’s a possible-to-memorize approach, with slightly better math, which allows for misses, variable success, advantage/disadvantage, and crits.

**Whenever a group of identical creatures make attack rolls** (or any roll really – you could profitably use this for group saving throws too), **make a single roll as normal. Divide the creatures into three roughly equal groups. One group rolls this number, one group rolls this with a +5 bonus, and one group rolls this with -5 penalty.**

**Implications of this system:** Advantage/disadvantage doesn’t require any special rules. Just make a single roll with adv/disadv and apply the group modifiers to the result. Auto-miss and crits work as you’d expect too. Because each group uses the same natural die roll, a natural 1 means everyone misses and 20 means everyone crits. That’s fun: the 20 hobgoblins do 40d8+20 (200) damage!

**The math:** What’s a better model of making 20 attack rolls: this system or the DMG system? Both are pretty good, actually, but mine exactly matches in most situations (whenever you need to roll a 6 to 16 to hit) while the DMG system is better at modeling corner cases (you need to roll a natural 20 to hit or you only miss on a 1). To me that’s not a big deal, because with bounded accuracy, even a bunch of town guards (+3) only need a 16 to hit an adult red dragon (AC 19).

Here’s a chart that compares the average results.

**Chance to hit per attack**

d20 roll needed | Rolling all attacks | DMG mob system | blogofholding mob system |

2 | 95% | 100% | 88.33% |

3 | 90% | 100% | 85% |

4 | 85% | 100% | 81.67% |

5 | 80% | 100% | 78.33% |

6 | 75% | 50% | 75% |

7 | 70% | 50% | 70% |

8 | 65% | 50% | 65% |

9 | 60% | 50% | 60% |

10 | 55% | 50% | 55% |

11 | 50% | 50% | 50% |

12 | 45% | 50% | 45% |

13 | 40% | 30% | 40% |

14 | 35% | 30% | 35% |

15 | 30% | 25% | 30% |

16 | 25% | 25% | 25% |

17 | 20% | 20% | 21.67% |

18 | 15% | 20% | 18.33% |

19 | 10% | 10% | 15% |

20 | 5% | 5% | 11.67% |

So that’s the system: three groups with +5, +0 and -5 modifiers! Go forth and drown your PCs with armies!

Awesome! I will try this out tonight at the table. The PCs are being ambushed by 40 bandits, and 15 or so caravan guards on their side.

Question: how would you handle critical hits with this system? My 40 bandits should average 2 critical hits per full set of rolls. So for them, I might do this:

Roll of 20: 4 crits

Roll of 16-19: 3 crits

Roll of 5-15: 2 crits

Roll of 2-4: 1 crit

Roll of 1: 0 crits

Or, to make a universal system, since each bandit has a 5% chance of critting, i could multiply 5% x 40 to get 200%, and then roll percentile dice as many times as possible until the added results exceeed 200. That’s a rolemastery solution though…but if you make the players make the rolls, it ends up being pretty fun at the table.

Thoughts?

I think that a natural 20 means every bandit crits. That’s very spiky but it doesn’t require any rules tweaking, and you can justify it by saying that, in mass combat, a critical hit represents a devastating attack from an advantageous position, and you just need to flavor it appropriately (charging downhill, turning the flank, firing when they see the whites of their eyes).

In my opinion, more tricky than crits will be adjudicating damage between groups, like the bandits shooting at the caravan guards. if 26 bandits hit, doing 5 damage each, how many 11-HP guards do they kill? I guess the easiest thing to do is to have the bandits focus fire: Each guard takes 3 hits to kill, so you tap the guards, “One two three dead, four five six dead,” until you’ve used up the 26 hits (killing 8 guards and bringing the 9th down to 1 HP).

I guess this is where context and story come into play. My PCs have been with most of the guards for about 10 sessions now, and they have become very fond of some of them. All of the guards use “named character” rules when they are reduced to 0 hp — they get to make death saves (which I have the players roll, of course). I know what will happen is that the combat-weak cleric of Lathander will be rushing around trying to stabilize them. Plus, the PCs will draw the most fire, as the bandits already know who they are and that they are really dangerous.

Long story short, rather than letting the dice fall where they may, in this case I will probably fudge the decision about who the bandits hit with an eye toward story — forcing the cleric to choose between stablizing different guards will be nicely dramatic and gut-wrenching.

Anyway — don’t mean to hijack discussion of your very useful system with my particular case! Carry on! =)

I’m using this system next game. Simple, elegant, and full of variety. Well done!

Brief update: On the very first turn of the bandit ambush last night, I rolled a 1, which meant that all of the bandits missed. Since there was a heavy rain, I improvised that there was a flash of lightning at the exact moment they fired, but I was not pleased with this result — it was the bandits best chance to do some real damage before the caravan had a chance to take cover, and while the fight that followed was a bit of a struggle for the PCs, it was never life and death.

One of my players suggested as an alternative to the “roll and add +5 and -5” method: Roll 5 d20s and drop the highest and lowest results. Or heck, just roll 3 d20s. I might try that next time (but in truth, the party has proven so effective at fighting hordes of weak critters that I might just have them fight dragons from now on. It’s the name of the game, after all).

I ran a session last night with smaller battles – the PCs vs 25 ghouls, followed by the PCs vs. a marilith and 15 ghasts. (I could have used this system for the marilith’s 6 identical attacks but I didn’t!) The ghasts never critically hit or fumbled, so I didn’t get to test that. Other than crits or fumbles, though, it worked pretty well for the ghasts’ attack rolls and saving throws: things kept cooking along even when the wizard was dropping fireballs on all the ghouls or all the ghasts were chewing on the wizard’s intestines.

Yes, rolling 3d20 for each group is a good alternative, and one that I considered – for me the +5/-5 way is a little quicker, especially with advantage, so I chose that, but another DM might find it quicker to read three dice, and the three-dice method does distribute the crits and misses nicely.

Something just occurred to me: maybe hits and fumbles don’t require any special effects at all for mass combat.

Normal combat has two sources of damage variation – a damage roll range, and crit damage is essentially a “degree of success” reward for rolling high.

boh.com mass combat has three sources of variation – a damage roll range, plus a “degree of success” reward for rolling high – more people hit – plus a second “degree of success” reward – extra crit damage for everyone. Maybe that’s double dipping.

I might change it to say that mass combat attacks ignore the auto-fumble or crit damage rules (although they do honor any specific special abilities that trigger on a crit).

Jake, even with rolling 3 dice, or 5 drop lowest and highest, you’d still have a 15-20% chance of every bandit missing on round one. The only way to guarantee some hits are to roll 40 dice (vanishingly small chance of all missing, but tedious) or using the DMG system (invariable 12 hits).

Paul, what about this simple method to handle the crits: as I’ve interpreted your rolling solution, the goal is that one third chunk of the mass has moderately lesser chance for succes, and on third has moderately better than the mean.

But nat 20-s and nat 1-s mean unexpected events, at least I think so.

So for a nat. 20 roll, I would decide, that one third rolled 15, so didn’t crit, and one third rolled 20, but not natural 20, so hit, but didn’t crit, and on third made the real nat 20.

For a nat. 1 roll, I’d take it as a one third nat. 1, on third not natural 1 (so minimal chance, for hit, but not automatic failura, and one third roll 6 (so slim chance to hit, but still chance, depends on attack bonus).

I just split the big group into small groups and roll separately. A swarm of 20 rats gets a pair of 7-rat attacks and a 6-rat attack, or four 5-rat attacks. But the swarms aren’t special monsters; a magic missile still kills one 1-hp rat, and a 1E 13th level Fighter with two attacks gets to hit two rats (although his sweep ability should let him attack one monster of less than 1 HD per Fighter level – I’d also have him roll a 6-rat attack and a 7-rat attack instead of all-or-nothing or 13 separate ones). A 3E Fighter with Whirlwind Attack hitting 20 rats would work a lot like 20 rats attacking that Fighter.

Works for mass combat too. Crits and fumbles are not a problem, it’s just how you describe them.

Rather than a misfortune befalling a person, a group suffers more from miscommunication and confusion. 20 bandits firing into a caravan could fail because the wind rose up suddenly and scattered their arrows, or someone in the caravan shouted in a way that echoed up and the bandits thought it was their leader calling for retreat (and in the confusion, they didn’t fire effectively). Maybe their footing was bad on their ambush point and they hadn’t noticed until they sprang forth. Maybe there was some astonishing or humorous event that drew most of the caravan people onto the other side of the caravan at exactly the right moment, so when the attack was called there were no good targets.

Similarly, when an attacking group thunders down upon an unready or distracted target, or one which suffers a momentary fright, they could inflict lots of extra casualties that you wouldn’t expect. And, assuming the damage is at least half from the roll instead of mostly a static bonus, 1/4 of the crits will cause damage within the normal hit range. For example, an arrow normally doing 1d6+2 damage will do 3 to 8, which means a crit damage roll of 1 or 2 will deal damage that could have come from a non-crit hit. Barring rules like 3E bows doing triple crits, or special crit effects, 1/3 of the crits from the arrow I described are indistinguishable to the enemy from normal hits. These, then, require no special description.

Here’s a simple way to deal with disadvantage and crits:

Advantage/Disadvantage- You still roll for damage, so roll that twice.

Crits- Damage dice explode. Gain 1 more die per maximum number.

Crits + Advantage- You only roll the “first” die roll twice, not the “extras”

That last rule is needed to keep crits from going too out of control. Big dice explode less, but then again, big weapons’ average damage will be leagues above smaller weapons, so I think it balances out.